PK Implements Custom Testing Framework for Colorado Non-profit
The client partners with public and private agencies to craft public policy and administer access to programs nationwide. Based in Colorado, the client is a national leader in cost-effective administration of public programs that include public health insurance, child support enforcement, and workforce development.
With a focus on implementing child support and other public programs across the country, the client’s business application had to accommodate the differing business rules in every state. The client’s software was workflow-driven, included third-party application programming interfaces (APIs), and incorporated strict federal and state rules that changed constantly. In order to maintain the system, the client had a number of complicated test cases which the in-house technical staff applied using separate, disjointed, time-consuming tools for test automation and performance testing. The client engaged PK to develop a framework for automation and performance testing that would help them achieve their business goals of:
- Reducing automation and performance testing costs.
- Improving the quality of their software.
- Reducing time-to-market for software releases.
- Incorporating changing business rules from the federal and state agencies to support the new bill passed by
President Obama in February 2009, which expanded the State Children’s Health Insurance Program (SCHIP) by roughly $35 billion over the following five years to provide federally funded healthcare to an estimated additional 4 million children.
Following the evaluation, PK customized the testing framework and applied additional logic to suit the client’s requirements for faster time-to-market and varying business rules from state to state.
PK created a time-boxed environment for tests, running test scripts as needed instead of allowing time for processes that were beyond their control. For example, when a user created a login to request child support, the verification process would take three days. To bypass this manual delay, PK would reset the date and run the automation test scripts.
PK customized their own Automation Testing Framework to meet the client’s requirements, using ApacheJMeter and Hudson for automation testing, performance testing, and continuous integration. The solution was implemented in a Linux environment and used approximately 500 test cases.
PK supported the client’s technical staff in development, testing, and implementation, and provided coaching in Agile methodology throughout the engagement.